AIN432 REPORT

Name-Surname = Hüseyin Yiğit Ülker Student Id = 21993092

The purpose of this assignment is to view the images of an image in different colour spaces, to obtain a colorful and high quality image by combining it with image processing techniques.

I followed these steps while doing the homework:

1- The process of reading images from the data set

2- Destroying the black and white areas on the edges of the read images

3- Separating the images from the horizontal black areas in the picture that separate the channels from each other

4- Common region scanning to merge images of different sizes that I separated as B, G, R

5- Image enhancement processes

Reading Input From Dataset

The data we use in the homework, as seen below, is taken with a single image containing 3 different images that correspond the blue, green and red channels of the same image.

Ekran Görüntüsü (42).png

Destroy Black and White Areas on the Edge

I cleaned up the unnecessary black and white edges before separating the pictures. While doing these operations, the image I dealt with the number of changes in pixel numbers in certain regions. I cleared pixels with fewer changes.

Ekran Görüntüsü (44).png

Seperate Each Image from the Horizontal Black Areas

While dividing the pictures into 3 parts, I determined the black line separating the pictures from each other and separated them from that region. While locating that region, assuming that 1/3 and 2/3 of the picture's height passes through the points, I chose the row with the most black pixels in 40 rows around these points as 20 below and 20 above. I set the black color limit to 40.

Ekran Görüntüsü (46).png

Ekran Görüntüsü (47).png

Ekran Görüntüsü (72).png

Common Region Scanning to Combine Images

I used the merge method to combine the pictures I divided into color channels. I used Gabor edge detection to find common points in all three images while applying the merging process. By scanning the rows 10 down and 10 up from the midpoints of the 3 filtered images, I combined them from the region with the least mean square error in each image.

Ekran Görüntüsü (73).png

Image Enhancement

I applied image sharpening as image enhancement so that the images I created are clearer.

Ekran Görüntüsü (74).png

ALL RESULTS

Image 1

Edited0sharp.jpg

Image 2

Edited1sharp.jpg

Image 3

Edited2sharp.jpg

Image 4

Edited3sharp.jpg

Image 5

Edited4sharp.jpg

Image 6

Edited5sharp.jpg

Image 7

Edited6sharp.jpg

Image 8

Edited7sharp.jpg

Image 9

Edited8sharp.jpg

Image 10

Edited9sharp.jpg

Image 11

Edited10sharp.jpg

Image 12

Edited11sharp.jpg

Image 13

Edited12sharp.jpg

Image 14

Edited13sharp.jpg

Image 15

Edited14sharp.jpg

Image 16

Edited15sharp.jpg

Image 17

Edited16sharp.jpg

Image 18

Edited17sharp.jpg

Image 19

Edited18sharp.jpg

Image 20

Edited19sharp.jpg

Image 21

Edited20sharp.jpg

Image 22

Edited21sharp.jpg

Image 23

Edited22sharp.jpg

Image 24

Edited23sharp.jpg

Image 25

Edited24sharp.jpg

Image 26

Edited25sharp.jpg

Image 27

Edited26sharp.jpg

Image 28

Edited27sharp.jpg

Image 29

Edited28sharp.jpg